Pilates: one small trial, in childrenMixed evidence
Our searches on 2 November 2026 found no Pilates trial in autistic people in PubMed or Europe PMC. One randomised trial, in a Turkish university journal, gave 22 autistic children aged 7 to 12 one 40-minute session a week for 12 weeks, of Pilates with a physiotherapist or treadmill and bike training; 21 were analysed. One-leg balance and ball skills improved in the Pilates group, but "all included outcome measures were statistically similar" between the groups, and social skills didn't improve in either. The aerobic group also got stronger.
Sources 3, 4
Why that trial is weakMixed evidence
It was tiny, the journal isn't indexed in MEDLINE, and it had no group that did nothing, so some gains may be practice or growing up. The groups differed at the start, and it didn't report side effects. Its registry record was submitted after the trial ended and gives 2021 dates, while the paper says 2017 to 2018. It found Pilates no better than aerobic training.
Sources 4, 5
Other Pilates reportsMixed evidence
A single-case study described reformer Pilates with a parent for one autistic 18-year-old, through interviews only, and a 2016 study's participant numbers don't add up. Neither can show whether Pilates helps.
Sources 21, 22
Exercise in general: movement skills and balanceEmerging evidence
In 10 trials (256 children, average age about 7), exercise improved overall movement skills by a large amount, with signs that small positive trials were more likely to be published. In 12 trials, balance improved, but programmes of 8 weeks or less made no clear difference. An earlier review of 29 studies (1,009 young people) found gains in movement skills, fitness, strength and social functioning, and a review of 41 studies found autistic children learned movement skills better from visual than spoken instructions, on low-quality evidence.
Sources 6, 7, 23, 24
Social skills and behaviour: uncertainMixed evidence
Across 11 trials, a small effect on social skills "was attenuated and no longer significant" once one study was left out, and an umbrella review that reanalysed the trials found no clear effect on social communication. Another umbrella review says "the overall certainty of evidence is low", and effects on core features of autism didn't last after programmes ended. Much of this research treats autistic traits, including stimming, as things to reduce, an aim the National Autistic Society doesn't share for harmless stims.
Sources 25, 26, 27, 28, 12
How sure we can beStrong evidence
Across 34 meta-analyses, "intervention effects increased progressively with publication year", a sign that the pooled numbers are inflated. About 1 in 4 children dropped out of exercise trials, no more than in comparison groups, and most children in the trials were boys aged about 5 to 13.
Sources 8, 29
Anxiety, sleep and thinking skillsEmerging evidence
Exercise reduced anxiety in 12 trials (482 autistic children), improved parent-reported sleep, and helped thinking skills such as planning and attention in 16 trials. All were small, mostly unblinded and in children, and none tested Pilates.
Sources 30, 31, 32
Yoga and other mind-body practicesMixed evidence
A review calls the evidence for yoga and mindfulness in young autistic people "inconclusive", and of 16 mind-body studies, only 3 were randomised trials. The largest yoga trial, in 43 children, improved whole-body but not fine movement skills. A 2026 network meta-analysis ranked mind-body exercise best for movement skills, but on 4 small studies, none of them Pilates.
Sources 33, 34, 35, 36
Autistic adultsEmerging evidence
A review of 22 studies in 763 autistic adults aged 19 to 30 found the strongest evidence for fitness, then movement skills, wellbeing and quality of life. We found no Pilates trial and no randomised yoga trial in autistic adults. The nearest is an adapted mindfulness course with gentle yoga (below).
Sources 37, 3, 38, 39
How Pilates compares
In the one head-to-head trial, Pilates and aerobic training did about equally for balance and movement skills, and the aerobic group also got stronger in 5 of 6 muscle groups, against 1 in the Pilates group. Pilates offers a fixed order, a slow pace and one-to-one teaching, but it isn't aerobic exercise, so it belongs beside active play, sport, swimming or walking (our reading).
Source 4